!

SPE 163003

Minimum Shut-in Time Prediction Model for Pressure Build-up Survey

for a Niger Delta Onshore Field
*Ekeoma Ejere, SPDC & Omeke James Emeka, SPE

Copyright 2012, Society of Petroleum Engineers

This paper was prepared for presentation at the Nigeria Annual International Conference and Exhibition held in Abuja, Nigeria, 68 August 2012.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have
not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of
Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers
is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of SPE copyright.

Abstract state during build-up; hence, it becomes

necessary to predict the minimum time of
Allowing a well to stabilise is an important part of stabilization in the presence of these complexities
both pressure drawdown and build-up test.If so as to minimize cost. This was the focus of this
stabilization is overlooked or is impossible, paper.
standard data analysis techniques may provide
erroneous information about the formation. One of The general formulas used in analyzing pressure
the principal objectives of pressure build-up build-up data came from a solution of the
analysis is to determine the static reservoir diffusivity equation with homogenous, isotropic
pressure without waiting weeks for the pressure in and slightly compressible single phase fluid
the entire reservoir to stabilize. The fact remains conditions. This equation is in its simplest form
that no matter how we try to avoid long shut-in and due to the assumptions made; it is obvious
period, the shut-in time must exceed the transient that the major uncertainties that may affect the
and well bore storage period if a reliable result is shut in time is not captured; hence a need for a
sought. The bulk of the analysis is done at the robust modelling.
middle transient period (MTP) which, sometimes,
due to the complex nature of the reservoir, may In this paper, a semi-analytical model for the
be appreciable enough to invalidate the intention Niger Delta onshore field was developed for
of shutting-in the well for just a short time. A optimum shut in time prediction for each well in
minimum shut in time of six hours has been a the reservoir selected for study. A correction
common conventional practice in our land function which encapsulates the real reservoir
locations. This brings us to the question - is this complexities was introduced based on the well
shut-in period enough or too small to provide a data under study. This correction function was
reliable analysis? This shut-in period remains a modelled using Gompertz, 4 parameter Sigmoidal
paradox as long as complexities and uncertainties equation1 with the correlating constant determined
are present in the reservoir under study. Some by the method of least square for the specific well
reservoir with varying degree of complexities has considered. A constrained nonlinear optimisation
varying shut in time. For example, the presence of technique was applied to the newly developed
faults and complex barriers may prolong the time model which serves as the objective function with
of the reservoir to reach its middle or late transient the aim of minimizing the shut-in time as well as

!
2 SPE 163003
attaining an effective stabilization pressure. This
procedure will, in no doubt, improve our optimism
of accurate shut-in time prediction, a cost effective
approach.
Introduction
Pressure transient analysis is performed in wells,
routinely or under special circumstances, to
determine an array of reservoir parameters to be
used in reserve estimation, performance
calculation and numerical simulation studies. The
simplest test involves measurement of pressure
either subsurface or at the tubing head over a
period of time. Analysis is performed on the
resulting data obtained by fitting the data to a
mathematical model which describes the fluid flow
in the reservoir. Once a good match is obtained,
the mathematical model will be used to calculate
the reservoir parameters. Types of well test
commonly performed includes: drawdown or
pressure decline test, pressure build-up test,
falloff test, infectivity test, drill stem test, etc.
In an ideal pressure build-up test situation, we
assume that the test is conducted in an infinite-
acting reservoir in which no boundary effects are
felt during the entire flow and later shut-in period.
The reservoir is considered homogeneous with a
slightly compressible, single-phase fluid with
uniform properties so that the Ei function and its
2
logarithmic approximation apply. Horners
approximation is applicable. Wellbore damage
and stimulation is concentrated in a skin of Zero
thickness at the wellbore. Flow into the wellbore
ceases immediately at shut-in. In actual build-up
tests in an infinite reservoir, instead of a single
straight line for all times, we obtain a curve with a
complicated shape, which indicates the effect of
afterflow, making it possible to divide a build-up
3
curve into three regions :
Early-time region In this region, a
pressure transient is moving through the
formation nearest the wellbore.
Middle-time region In this region, the
pressure transient has moved away from
the wellbore into the bulk formation
!
Late-time region In this region, the
pressure transient has reached the
drainage boundaries of the well
Middle time region is a straight line which
conforms to Honers equation. This is the portion
of the build-up curve that must be identified and
analysed. Analysis of this portion only will provide
reliable reservoir properties of the tested well. The
distortion of the straight line at the early and late
time region is due to altered permeability near
wellbore, wellbore storage and effect of reservoir
boundary on flow respectively.
Major analysis and time models developed by
most authors are focused on the flow time before
shut-in. They were derived based on the flow
regimes as the well tends to stabilization when the
fluid front reaches the external boundaries. For
4 5
example, Horn and Earlougher among other
authors have indicated that the wellbore pressure
is directly proportional to the time during the
wellbore storage-dominated period of the test.
5
Earlougher pointed out that the time for the end
of well bore storage can be estimated from the
following expressions:
!""!!!! ! !"###! !
!!
!!!!
Where
! = skin factor, ! = wellbore storage coefficient,
bbl/psi, !!= product of permeadility and thickness,
Md and ft and != viscosity.
The time taking to reach a pseudo steady state is
given as
!"#!"!!" !
!!"" !
!
Where
!!" = initial total compressibility /psi, != porosity,
2
and A = area ft
A relationship between the short in time of a well
in the presence of sealing fault is given by the
3
following relationship
3 SPE 163003
!"####!"!! !!
!! !
! purposes, this implies that all three
Where !!= minimum short in time, hr. And L =
distance between well and the sealing fault, ft.
7-9
Different authors has written substantial time
model for a flowing well with different flow regime.
They each expressed time to reach the early
transient, middle transient and late transient or
pseudo-steady flow regime for horizontal wells.
In as much as efforts made by previous authors
were mostly focused on predicting time to attain
some flow regimes under flowing condition, little
or no consideration was made in developing,
explicitly, the minimum time required to shut in a
well. Most of these time models where expressed
as inequalities due to uncertainties involved. Also,
their result may be misleading since they were
based on the assumption imposed on the host
equation from which they were derived. Minimum
shut in time prediction is very important for
accuracy in reservoir properties estimation as well
as economic consideration. This was achieved in
this work.
Assumptions involved in traditional well test
equation
In the process of developing the traditional
pressure transient applications, the following
assumptions were generally taking into
consideration:
Single phase flow of either oil or gas
1D (one dimensional) radial form of oil
flow equation
There are no external sources or sinks,
when compared with the numerical
simulation equation, which implies no
well. Qsc = 0
The reservoir is horizontal, which implies
that
!!
!
!"
the first term represents the
!" !"
derivative of the low potential while the
second term represent the derivative of
pressure, all with respect to radius
The effective permeability to oil is
independent of position (that is, a
!
homogeneous distribution). For practical
components (irreducible water saturation,
absolute permeability and relative
permeability characteristics ) are
independent of position
The viscosity is constant
The pressure gradient is small
The oil-phase compressibility is small
With these assumptions, the pressure build-up
equation was introduced by Horner and is
commonly referred to as the Honer equation. It is
given as follows:
!"#!!!! !! !! !! ! !!
!!" ! !! ! !"# !!!
!! !!
Where:
!! = initial reservoir pressure, psi
!!" = sand face pressure during pressure build-
up, psi
!! = flowing time before shut-in, hours
!! = stabilized well flow rate before shut-in,
STB/day
!! = shut-in time, hours
The above equation shall be used in course of the
development of the new model
New model development/approach
Equation 1 is indeed an ideal representation of
the reservoir condition. The assumption assigned
to it limits its accuracy when compared to real
field case. For example, if all the reservoir data
required by the Horner equation was substituted
into the equation, all data gotten from one field,
the prediction of !!" when compared to that read
from the real field measurement yields values that
are up to 20%-90% difference. The difference
results from the idealization of the reservoir
condition which are reflected in these equations.
The basis of our formulation is to adjust this ideal
state to a real state by introducing some adjusting
 4 SPE 163003

functions. The following are the processes Procedures involved in determining the
involved: constants

Lets define the following terms as follows: The model above is a nonlinear model and hence
requires a nonlinear regression analysis. The
!"#!!!! !! !! !! ! !! following procedure was carried out for the
!!"!!"#$%! ! !! ! !"# !!
!! !! !"#$% optimization of the fitting parameters that best
describes the dependent and the independent
Where
variables with a well more than (M+1)-point field
!!!"!!"#$%! data set.
! !"#$%!!"#$%"&'!!"#$$%"#!!"#$%&!!"#$%"&! !"#
1. An Initial guess of !! !! ! !! !!"#!!!was performed
2. Calculate !!"!!"#$%! using Horners equation
And
!!"!!"#$!
3. Calculate the ratio by dividing the measured
!!"!!"#$%!
!!"!!"#$!
field sand-face pressure with that calculated from
! !"#$%"&'!!"#$$%"#!!"#$%&"'!!"#$!!"#$%! !"#
Horners equation
Therefore the correction to the ideal case can be 4. Calculate a new variable
!!!!!
performed empirically by the following relationship !
!
! !! ! !!!! that applies the
!
!!"!!"#$! ! !!"!!"#$%! !!! ! ! !! ! ! !! assumed fitting parameters to every data point
!!!!!
!
Where (i) of the field data set. !! ! !! !! !

!!!!!
!
!! ! ! !! ! !! ! !!!!
!
! ! !! must have similar trend or close resemblance to
!!"!!"#$!
the trend generated by the ratio
!!"!!"#$%!
The above equation is a Sigmodial Hill, 3
5. Calculate the squared error !!!
parameters equation. The trend generated by this
objective function for every data point as
equation closely approximated the trend that exist
shown below.
in a build up test starting from the early transient !!!!!
!
period to the late pseudo-steady state. The !"! ! !! ! !!!! !

parameters !! !! ! !! !!"#!!!will be determined !"#$!!

!
empirically using least square approach. !!"!!"#$!
!!
!!"!!"#$%!
The form of the new developed model will be !"#!!
represented as follows 6. Add the objective function to obtain the
sum-squared error (SSE):
!"#!!!! !! !! !! ! !! !!"
!!"!!"#$! ! !! ! !"#
!! !! !"#$%
!
!!!!! ! !!! !!!!!!!!"!!!!!"!!"#$%&!!"!!"#"!!"#$%&
!
!! !
! !! ! !! ! !! !!!
7. Apply an optimization subroutine (e.g
This can also be expressed as Microsoft Excel Solver) for minimizing the
SSE by changing the fitting parameters.
!!!!! The final fitting parameters that result in
!!"!!"#$! !
!
! !! ! !!!! ! ! !! the least SSE (LSSE) are the final result
!!"!!"#$%!
for the linear regression
8. The average relative error (ARE) for the
obtained correlation can be calculated as:

!
5 SPE 163003

!"#!
!!!!! ./00%1%!2332*%,24+%&-+''5-+%,('32-6%
!!"!!"#$! !
!
! !!"!!"#$%!
! !! ! !!!!
!"" !"#!! !"#$!!
!
! !!"!!"#$!
!!!
!!"!!"#$%!
!"#!!

!"#$%&'(%
Optimization procedure

The objective of this work is to estimate the

minimum shut in time required as the reservoir
pressure tend to a stabilized value. In other to
achieve this, the following procedures were used:
)(*+$%,-'%

Allocate a default or initial values as

inputs to equation 5 with equation 5 as
Fig 1. BHP History plot for WELL A
the Objective function
Invoke the Microsoft excel solver to Below is a table of the reservoir and fluid data
perform the optimization, setting the used for the analysis.
following condition:

!!!"!!"#$%!!"!!!!!"##!!"#$%&!!"!!"#$%$&$'(!
! !!!"!!"#$!
!"#!!!! !! !! !! ! !!
! !! ! !"# Table 1. Data Table for WELL A
!! !! !"#$%
!!!!!
!
! Reservoir properties for
! !! ! !!!! WELL A
!!"!!!!"!!"#$! ! ! !! !!! Qo(Stb/d) 488
o (cp) 0.8
From the above expression, as the sand-
Bo(rb/Stb) 1.253
face pressure build-up to its average
K(Md) 830
reservoir pressure values, the value of the
!(porosity) 0.28
Shut-in time also approaches its optimum
tp(hr) 0.5
value
P (psia) 2780
The following constrain was taken into
h(ft) 88
consideration
!"#!!!"!!!"#! ! !! !! !!
!"#!!"!!"!!"#"$%&%#'!!"#$%&#%!!"#!$%!!! Result for example 1

Application After performing the nonlinear regression on

WELL A field build-up shut-in time data, the
Field example 1
following coefficients where obtained from the
expression:
The needed data were extracted from the
information provided for Well A in reservoir X, in
yo 0.236
Niger Delta onshore field in. The field is an
a 0.778
onshore field. Fig.1 below is a plot of the bottom
xo 8.175
hole pressure history of the well.
b 0.553

!
 6 SPE 163003

Therefore, for WELL A in reservoir X, the Field example 2

following holds:
The needed data were extracted from the
!!"!!"#$! !
!!!!!
!
information provided for Well B, reservoir Y, in
! ! !! ! !!!! same Niger Delta onshore field Fig.2 below is a
!!"!!"#$%!
!
!!!!!!"# plot of the bottom hole pressure history of the
!!!!"
! !!!"# ! !!!!"!!!! well.

The above equation can also be expressed as

follows:

!!"!!"#$! !2332*%,24+%&-+''5-+%,('32-6%?2-%./00%!%
! !!!"#
!!"!!"#$%!
! !!!!"
!! ! !!!"#
!!"# !!"# !
!!!!"

!"#$%&'7%
In its full form, the equation for WELL A can be
expressed as

!"#!!!! !! !! !! ! !!
!!"!!"#$! ! !! ! !"#
!! !! !"#$%

! !!!!"#
)(*+$%,-%
! !!!!"
!! ! !!!"#
!!"# !!"! ! ! !!
!!!!" Fig 2. BHP History plot for WELL B
Below is a plot of the matching information for Table 2. Data Table for WELL B
WELL A
Reservoir properties for
WELL A
:73;,(<=%./00%1%>(3,%:29+4%
Qo(Stb/d) 224
o (cp) 7.9
Bo(rb/Stb) 1.13
K(Md) 997
&-+''5-+%-73(2%

!(porosity) 0.28
tp(hr) 0.5
P (psia) 2827
h(ft) 86

Result for example 1

85(49%5&%3(*+$%,-%
!"#$%& '(%#$&
With

yo 0.142
The average absolute value of the relative
a 0.809
error is given as ARE = 1.4 %
xo 8.723
b 0.188

!
7 SPE 163003

The following model was also obtained from study is small, the general correlation
WELLB constant that will be applicable to Niger Delta
onshore field and the proposed model is given
!!"!!"#$!
below
!"#!!!! !! !! !! ! !!
! !! ! !"#
!! !! !"#$%
AVG
! !!!!"# yo 0.235885 0.142374 0.140137 0.172798
!! ! !!!"# a 0.777988 0.809064 0.815727 0.800926
! !!!"#!!"# !!"# ! !!
!!!"" xo 8.175314 8.722966 8.722842 8.540374
With ARE = 1.3% b 0.552933 0.187536 0.190957 0.310475

Field example 3
The final form of the Niger Delta onshore field
The needed data were extracted from the model is thus
information provided for Well C, in same reservoir
Y, in same Niger Delta onshore field Fig.3 below !"#!!!! !! !! !! ! !!
!!"!!"#$! ! !! ! !"#
is a plot of the bottom hole pressure history of the !! !! !"#$%
well.
! !!!!"#
Result for example 3 ! !!!"#
!! ! !!!"
With !!"# !!"# !
!!!"

yo 0.140137 After using the above equation for the

a 0.815727 analysis of each well in other to predict each
xo 8.722842 maximum and minimum shut-in time, the
b 0.190957 following result where obtained:

WELL A Maximum shut-in time = 15 hrs

The following equation was obtained for Well and Minimum is = 10 hrs
C
WELL B Maximum shut-in time = 15 hrs
!"#!!!! !! !! !! ! !!
!!"!!"#$! ! !! ! !"# and Minimum is = 9.5 hrs
!! !! !"#$%

! !!!!"# WELL C Maximum shut-in time = 11 hrs

! !!!"# and Minimum is = 9 hrs
!! ! !!!"#
!!"# !!"# ! Conclusion
!!!"

With ARE = 1.1% An approximate correlation has been

Discussion
It was observed that the correlation
coefficient gotten for all the Wells under study
where similar, except for a slight difference.
Since the difference between the values of
!! !! ! !! !!"#!! for each well and field under
developed that will enable the prediction of
minimum and maximum shut-in time in Niger
Delta onshore field prior to Well testing.
The model is formation based and will yield
accurate result only within the field studied.
Though this study is still at its preliminary
stage, the result gotten from the model is

!
8 SPE 163003
satisfactory when compared with the field
data used.
References
1. Sigma Plot 12.1 . www.sigmaplot.com
2. Horner, D.R. (1951). Pressure build-
up in wells. Proceedings of the Third
World Petroleum Congress, The
Hague, Sec II, 503523. Also
Pressure Analysis Methods, Reprint
Series,No. 9 (Dallas, TX: Society of
PetroleumEngineers of AIME), pp.
2543
3. Tarek, Ahmed and Paul. D.M (2005).
Advanced Reservoir Engineering
textbook,! Gulf Professional
Publishing.
4. Horn, R. (1995). Modern Test
Analysis, (Palo Alto, CA: Petroway).
5. Earlougher, Robert C., JR. (1977).
Advances in Well Test Analysis,
Monograph Vol. 5 (Dallas, TX: Society
of PetroleumEngineers of AIME)
6. Odeh, A. S., and Jones, L. G.,(1965)
"Pressure Drawdown Analysis -
Variable-Rate Case . Pet. Tech.
(Aug. 1965) 960-964; Trans. AIME, 23
7. Joshi, S. D.,(1988) "Horizontal Well
Production Forecasting Methods and
a Comparison of Horizontal Wells and
Stimulated Vertical Wells," paper
presented at the NPD Seminar on
Thin Oil Zone Development,
Stavanger, Norway, April 21-22, 1988.
8. Odeh, A. S., and Babu, D. K., (1990)
"Transient Flow Behavior of
Horizontal Wells Pressure Drawdown
and Buildup Analysis," SPE Formation
Eval.(March 1990), 7-15.